Saying a DSLR has 10 stops of dynamic range sounds impressive, but that still only amounts to a 1024:1 ratio, which is considerably less than many common real-world situations, and (as experience tells us) less than b&w film can capture with careful exposure and development.
And while a 10-bit or 12-bit film scanner in theory shouldn't be able to extract any more dynamic range from the film than a digital camera can extract from the original scene, in practice you can get more range out of the film for two reasons: (1) the film's contrast curve already has compressed the tonal information to some extent, and (2) if necessary, you can scan the film more than once at different settings to extract detail out of both the low-density and high-density areas.
This is roughly equivalent to making multiple bracketed exposures with a DSLR and then assembling them into an "HDRI" image. However, with film you don't have to worry about objects moving between exposures! Dye-based films generally have a low enough D-max that this trick isn't necessary, but when scanning conventional silver-based films I often scan them two or even three times to get good tonal distribution in the shadows, midtones, and highlights. I put the three scans in separate layers in Photoshop and adjust their opacity until I get a full-range image; I can also use layer masks to make localized corrections to areas that need to be lighter or darker, without the loss of detail that occurs when using Photoshop's 'dodge' and 'burn' tools.
I shoot almost 100% digital now for time reasons, but I often find myself shooting in places where I used to use film, and I've learned to accept highlight "blowout" under conditions that I know from experience I would have been able to "save" when shooting in the same type of situation on film.
